Overview

The performance of airlines is measured on a yearly basis and ranked from best to worst. Alaska Airlines was regarded as the best airline in America by Wallethub in 2019. On the same list, Frontier ended up in last place. During this project we will be analyzing a subset of this list and look at their 2018 performance, but also take a look into their history and see what we can find there.

We will be taking a look at five different airlines:

We chose these airlines as they fairly well spread out in the ranking list published by Wallethub, https://wallethub.com/edu/best-airlines/20916/. In addition, Frontier Airlines is a Low-Cost Carrier (LCC) which means that the airline puts high emphasis on minimizing operating costs and providing the cheapest tickets possible. This makes it unique to the other airlines which do not have this business model.

We will be analyzing each airline on a number of categories:

We have also looked at the performance of different American airports, especially regarding the On-Time performance and the number of flights in the year 2018.

Initial Question

The initial question that sparked this project is are we able to determine the best or worst airline in America? Just like the related work previously mentioned we will look at the different categories and determine good and bad airlines.

Over the course of this project we have also added in airport performance over 2018. The current implementation forms an introduction and could be used as a point of future work to further analysis the performance of certain airports.

Data

Since we have a variety of categories, we have used a number of different data sources and a number of different ways to process the data.

Customer Data

To gather the customer data we have scraped the review system of Flightradar24. Reviews can be found for every airline and have a number of categories:

  • Boarding / Deplaning (The process of getting on and off board of an aircraft)
  • Flight Crew
  • In-Flight Services
  • In-Flight Entertainment
  • WiFi

An example page for Alaska Airlines: https://www.flightradar24.com/data/airlines/as-asa/reviews

Flight Data

The flight data is gathered from the Office of Airline Information, Bureau of Transportation Statistics (BTS). The data ranges from 1987 till 2019 and every month comes in the form of a csv file. We have combined these csv files per year and used this to analyze the history. source: https://www.transtats.bts.gov/DatabaseInfo.asp?DB_ID=120&Link=0

Most csv files are large in size, ~500 Mb. Therefore we make use of fread instead of read.csv since it is way faster. For cancelled flights we decided to use relative values mainly because all airlines have a different number of total flights. Since a big airline will have more cancelled flights this makes it difficult to have an accurate comparison.

Financial Data

The Airline Data Project by the MIT has gathered financial data for most major airlines. We have used a portion of the data they provide to compare the financial state of airlines. The data comes in the form of excel files. source: http://web.mit.edu/airlinedata/www/default.html

Exploratory Analysis

All graphs are made with ggplot2 and plotly and fully interactive, you can however over data points to see the exact value and compare them with other airlines. You can make use of the bar below time plots to zoom in on specific time periods. You can do this by dragging the bar from the left and right towards the middle. You can also filter out airlines by clicking them in the legend. For the interactive airport map you can click on a marker to view more information about each airport.

Customer Reviews

Overall Customer Review Scores

As can be seen from the plot shown above, it is clear Alaska Airlines and Delta Air Lines have the highest overall customer reviews. On the lower end we find Frontier Airlines. It seems that Frontier customers fly at a lower cost, but this also impacts their satisfaction. Thus we might conclude that a cheaper flight is not always worth it. The difference becomes even more apparent if we take a look at the WiFi and In-Flight Entertainment reviews of all airlines.

Here we see that Frontier Airlines is by far the worst scoring airline with a review score of 34 for In-Flight Entertainment and 29 for WiFi. However these scores can be explained by the business philosphy that Frontier Airlines has adopted. They want to offer the cheapest flights possible and thus try to cut down on things such as In-Flight Entertainment and WiFi.

Delta and Alaska have the highest score. Delta offers free Wifi (https://www.delta.com/content/www/en_US/traveling-with-us/onboard-services/delta-studio/in-flight-wi-fi.html) and it seems to do them well in the review scores.

Boarding and Deplaning reviews

With Boarding and Deplaning we see a similar trend where Alaska Airlines and Delta Air Lines are at the top. Next to that we also see Frontier Airlines having the lowest score, again. We will look into the departure and arrival delays to see if we can spot the same trend. We might expect frontier to have more delay on average since that would justify the lower review scores.

We can see from the boxplot shown above that, on average, Frontier Airlines has more departure delay than other airlines. However the majority of departure delays are within 14 minutes of the scheduled departure time and thus the actual delay is relatively small. American Airlines, Delta Air Lines and United Airlines are more consistent in this category since their upper and lower fences are not as far apart, indicating a smaller spread in departure delays. Alaska Airlines is somewhere in between these groups.

Looking at the arrival delay we see a similar trend as to the departure delay with Frontier Airlines scoring the worst together. Interesting is that Alaska Airlines scores similar to Frontier Airlines in this category. All major airlines have a similar score.

Looking at the history

The graph above shows the number of flights per month and clearly visualizes the size difference between the major airlines (American, Delta and United) and Alaska and Frontier.

Some notable data points can be observed when looking at this graph. First of all is September 2001 and the impact it had on the number of flights for every airline. We see Delta Air Lines dropping from 74892 flights in August 2001 to 60133 in October 2001. Interesting to note is that it did not seem to impact Alaska Airlines as much as it did Delta, American and United.

Another interesting point is happens in January 2010. This is when Delta Air Lines has a significant increase in the number of flights per month. This is when Delta Air Lines and Northwest Airlines officially merged and the Northwest brand officially retired. source: https://www.reuters.com/article/us-delta-northwest/delta-buys-northwest-to-create-biggest-airline-idUSTRE49S8BA20081030

We see a similar thing happening to United Airlines in January 2012 who merged with Continental Airlines. source: https://www.nytimes.com/2010/05/03/business/03merger.html

We also see the exact point when American Airlines officially merged with US Airways, April 2015. source: https://www.washingtonpost.com/business/the-last-days-of-us-airways/2015/09/25/f5530686-60a6-11e5-8e9e-dce8a2a2a679_story.html

On a smaller scale we see the merge from Alaska Airlines and Virgin America in April 2018. source: https://blog.alaskaair.com/alaska-airlines/company-news/virgin-america-alaska-single-operating-certificate/

The most interesting data point is likely January 2000, where there is a sudden spike in the percentage of cancelled flights. It could be the case that the Y2k had influence and airlines preemptively cancelled flights, however we could not find a source on this. As far as the airlines are concerned, American Airlines had the most cancelled flights relatively speaking for a long time, but has been overtaken this year by Frontier Airlines. Frontier had a low amount of cancelled flights for a long time but the percentage has been rising for the last couple of years. At the lower end we see Delta Air Lines having a low amount of cancelled flights.

Looking at the average departure delay in the history we see that since 2010 Alaska Airlines has consistenly had the lowest amount of departure delay. On the other side of the spectrum we find Frontier Airlines, which consistently has had the most average departure delay. Interesting is that the average departure delay has increased since 2017. Whereas before it could be said that Frontier was comparable to American, Delta and United, it has since grown a gap to those other airlines.

As far as the average arrival delay, Alaska and Delta Airlines seem to be the best performing in 2018. Though it should be noted that the airlines seem to be more closely matched in this category. However Frontier is still the worst performing in this category. Also looking throughout the history, Frontier seems to be the airline with the most average arrival delay. Delta and Alaska have the least amount of average arrival delay throughout its history.

Cancellation reasons

From the plot shown above it is difficult to show that one airline is better than the other. On the other hand we do see a pattern across all airlines. During the winter period (December till March) we see that the majority of flight cancellations is due to the weather. This should not really come as a suprise, since bad weather that would influence airlines is more likely during the winter as opposed to the summer. This can be an interesting point to follow up on in future work.

Taking a look at the Total Full-Time Employee equivalent graph, we see a major disparity between the major airlines (American Airlines, Delta Air Lines and United Airlines) and Alaska Airlines and Frontier Airlines. This should not come as a suprise, but it makes a good indication of how much bigger the major airlines are. We also see each major airline increasing this number, Delta in 2010, United in 2012 and America in 2015. This directly corresponds to the merges that each airline made which have been mentioned in an earlier graph.

Over time we see that all airlines except Frontier Airlines have had similar annual wages and salaries. Only since 2015 has Frontier closed the wage and salary gap that has existed since the airline started. The highest average wages and salaries can be found at Delta Air Lines and United Airlines.

The graph above shows a fairly similar picture as the graph before. All airlines are comparable to each in terms of the average pension and benefit package, except Frontier who has been significantly behind since the beginning. Whereas Frontier has closed the wages and salaries gap with the other airlines, it does not seem likely to close the pension and benefit package gap in the foreseeable future.

ASM = Available Seat Mile, this means one aircraft seat flown one mile. This allows us to make comparisons between airlines that fly different aircrafts for different stage lenghts.

As can be seen from the graph above, Frontier Airlines has the highest score at 48,95 in 2018 and consistently scores higher than all other airlines. This could indicate that Frontier employees have a higher productivity for their pay rate compared to other airline employees. All other airlines show similar scores indicating equal productivity for their pay rate. This could be due to the fact that Frontier is a Low-Cost Carrier, which means they are more focused on producing more ASM per Dollar. Next to that they do not provide In-Flight Entertainment and Wifi which needs maintenance. The money they save with this might lead to more ASMs.

This last graph shows the number of employees per aircraft. Interesting to note is that since 1995 the average employees per aircraft has gone down*, except for United. It could be the case that United has an older fleet which requires requires more maintenance, howerver we cannot deduce that from this graph alone.

*We exclude Frontier since they were only a minor airline in 1995 with a small amount of employees.

Airport performance in 2018

On this interactive map you can view all American airports. Clicking on a marker will give more information about the airport, this includes On-Time performance and the number of flights in 2018. The color of a marker indicates their On-time performance, ranging from green (good) to red (bad).

Final Analysis

We started this project with the question whether we can find the best and worst airline in America. Looking at all the categories it becomes clear that Frontier Airlines is, with margin, the worst airlines out of all that were analyzed. Having the lowest score in almost every category.

American Airlines and United Airlines battle it out for third place and it is difficult to name one of these airlines as a clear winner. American Airlines scores better in customer reviews, but worse in the percentage of cancelled flights. United Airlines also outperforms American Airlines when it comes to the employee treatment, offering better wages and salaries and better pension and benefit packages over the last couple of years. So one could say that United Airlines takes third place, but only marginally.

The title of best airline goes between Alaska Airlines and Delta Air Lines. Both airlines have similar customer review scores, Alaska has a slightly better on-time performance, but relatively more cancelled flights. Delta offers a better average annual wage and salary. Both airlines offer a similar pension and benefit package. The airlines are closely matched and thus we will both name them ‘The best Airline of America’.